Superfamily CHALCIDOIDEA 745 



lay their eggs on the host's body and the larvae develop externally, or deposit their eggs inside 

 the host's body and the larvae develop internally. There is a tendency for parasites that attack 

 exposed hosts to develop internally (exception: elachertine Eulophidae), and parasites that at- 

 tack concealed hosts to develop externally. 



The intra- and interspecific relationships among parasitic chalcidoids vary. Some species are 

 solitary (one parasite per host), and others are gregarious (several parasites per host). When 

 more than one parasite species develops on a host simultaneously, the condition is termed multi- 

 ple parasitism. When more eggs of one parasite species are laid on a host than can develop to 

 maturity, the condition is termed superparasitism. Supernumerary individuals are eliminated 

 through larval combat or physiological suppression (Salt, 1961). 



The distinction between parasitism and predation sometimes fails, and some chalcidoids could 

 be called predators. A prime distinction between parasites and predators is that predators 

 frequently consume several prey, but parasites consume only one host per individual. The eu- 

 notine pteromalids and some mymarids could be regarded as egg predators because their larvae 

 feed externally on scale-insect eggs in the "brood chamber" after they are oviposited by the 

 female scale-insect (Clausen, 1940 a). 



Parasitic species that attack many species of hosts are called polyphagous; parasitic species 

 that attack only a few species of hosts are called stenophagous; and parasitic species that attack 

 only one species of host are monophagous. Complete host specificity is difficult to establish 

 because it is based essentially on negative evidence. The fact that a parasite will not attack a 

 host under some conditions does not constitute proof that it will not parasitize that species. 

 Nevertheless, there is a tendency towards specialization in the Chalcidoidea, and this is reflected 

 by: (1) repeated recovery of a parasite from a host species over a large area, but not from re- 

 lated host species that occur sympatrically; (2) demonstrated preference for a host species when 

 a choice is available; (3) superior reproductive capability on a host species; and (4) physical 

 limitations that prevent a parasite from attacking a potential host. 



Some polyphagous chalcidoids appear to prefer habitats rather than a taxonomically cohesive 

 group of hosts. For example, Zagravimosovia species parasitize leaf-mining insects whether 

 they are Lepidoptera, Diptera, or perhaps Hymenoptera. In contrast, related Diglyphus species 

 parasitize only leaf-mining agromyzid Diptera. Other chalcidoids are extremely polyphagous. 

 Dibrachys cavus (Walker) is an example. This species, like several others, has an exceedingly 

 long host list that includes representatives of several orders. It usually develops as a primary 

 parasite, but frequently also acts as a facultative hyperparasite (Graham, 1969). No explanation 

 has been provided as to why one species should be so polyphagous and a closely related, 

 morphologically similar species should be stenophagous or even monophagous. 



Likewise, there are associations between host stage attacked and the taxonomic assignment of 

 the parasite. For instance, the Trichogrammatidae and Mymaridae exclusively develop on the 

 egg stage of other insects and the spalangine pteromalids are pupal parasites (Annecke and 

 Doutt, 1961; Boucek, 1963; Doutt and Viggiani, 1968). 



Chalcidoids parasitize more hosts in more different taxonomic categories than any other 

 group of parasitic insects. This spectrum extends from spider eggs (Desantisca) to aculeate 

 Hymenoptera {Melittobia, Leucospidae). A detailed account of the biology of chalcidoids 

 requires more space than is available here. However, a short summary of some interesting host 

 relationships is provided. 



A bizarre host association is found in Ixodiphagus and Hunterellus (Encyrtidae) whose spe- 

 cies are primary, internal parasites of tick larvae and nymphs. These genera are cosmopolitan 

 and may prove to be beneficial insects in tick control (Cooley and Kohls, 1934; Cole, 1965; Doube 

 and Heath, 1975). 



The mymarid Caraphractus cinctus Walker is unusual in that it parasitizes dytiscid beetle 

 eggs that are submerged beneath the surface of the water. The female parasite swims in the 

 water by vibrating her wings and oviposits in the host's eggs. Females have considerable dis- 

 criminative ability, and can detect eggs that have been parasitized (Jackson, 1958, 1966). 



The Eucharitidae are parasitic on Formicidae. The association apparently is an old one, and 

 eucharitids oviposit on vegetation visited by worker ants. The eggs hatch, and the triungulin lar- 

 vae are phoretically transported to the ant nest. Inside the nest the triungulin larvae eventually 

 move into the brood chamber where they parasitize immature ants (Clausen, 1923; 1940 b,c). 



Other information about host association of chalcidoids is limited by a lack of knowledge about 

 the immature stages of many groups of potential hosts. However, the higher taxonomic catego- 



